High-Mileage Light-Duty Fleet Vehicle Emissions - ACS Publications

May 3, 2016 - involve age, initial emission certification levels, and emission level classifications (low to high). Both of these ... further investig...
0 downloads 16 Views 2MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article pubs.acs.org/est

High-Mileage Light-Duty Fleet Vehicle Emissions: Their Potentially Overlooked Importance Gary A. Bishop,*,† Donald H. Stedman,† Daniel A. Burgard,‡ and Oscar Atkinson‡ †

Department of Chemistry and Biochemistry, University of Denver, Denver, Colorado 80208, United States Department of Chemistry, University of Puget Sound, Tacoma, Washington 98416, United States



S Supporting Information *

ABSTRACT: State and local agencies in the United States use activity-based computer models to estimate mobile source emissions for inventories. These models generally assume that vehicle activity levels are uniform across all of the vehicle emission level classifications using the same age-adjusted travel fractions. Recent fuel-specific emission measurements from the SeaTac Airport, Los Angeles, and multi-year measurements in the Chicago area suggest that some high-mileage fleets are responsible for a disproportionate share of the fleet’s emissions. Hybrid taxis at the airport show large increases in carbon monoxide, hydrocarbon, and oxide of nitrogen emissions in their fourth year when compared to similar vehicles from the general population. Ammonia emissions from the airport shuttle vans indicate that catalyst reduction capability begins to wane after 5−6 years, 3 times faster than is observed in the general population, indicating accelerated aging. In Chicago, the observed, on-road taxi fleet also had significantly higher emissions and an emissions share that was more than double their fleet representation. When compounded by their expected higher than average mileage accumulation, we estimate that these small fleets (2−5% of fleet emissions).



INTRODUCTION State and local vehicle emission control measures in the United States depend upon the preparation of accurate mobile source emission inventories for evaluating effective emission reduction alternatives. Most inventory calculations rely upon a computer model to estimate the fleet’s emissions. In California, EMFAC is the model required for on-road emissions, and in the rest of the U.S., it is the United States Environmental Protection Agency’s (U.S. EPA) MOVES model.1,2 Both models in general calculate a mass of emissions emitted through the combination of fleet size, fleet average emissions, vehicle speeds (EMFAC), and fleet average mileage activity. Fleet size and activity are linked with age (model year), while mean emissions generally involve age, initial emission certification levels, and emission level classifications (low to high). Both of these models use different approaches to accomplish the end result, but in both models, vehicle activity levels are most often uniformly applied across all of the vehicle emission level classifications, with the underlying assumption that all vehicles, including high emitters, have the same age-adjusted travel fractions. Modern vehicle fleet emissions are dominated by a relatively small number of high emitting vehicles, resulting in a skewed emission distribution, which is best described by a gamma distribution.3 High emitters occur as a result of breakage, malmaintenance (performance modifications, rechipping, and/or after-treatment removal), and at end of a vehicle’s useful life. © 2016 American Chemical Society

For example, recent on-road emission measurements in California and other locations have shown that the 99th percentile is responsible for more than a third of the carbon monoxide (CO) and hydrocarbon (HC) emissions and more than 15% of the nitric oxide (NO) emissions on a fuel-specific basis, and these imbalances continue to grow each year.4,5 These percent contributions assume that the fuel consumption is proportional to the fleet fraction sampled. However, these percentages could increase or decrease if the vehicles represented in the 99th percentile consume significantly less or more fuel than the fleet average, potentially adding an additional compounding skewed distribution to the calculation. One segment of the on-road fleet that is believed to drive more miles and consume more fuel than the average fleet vehicle is livery (for-hire) vehicles, such as taxis and limousines. A recent news article about a New York City taxi company highlighted the fact that their taxis routinely accumulate 100 000 miles/year operating 7 days a week on two 10 h shifts each day.6 With such a rapid accumulation of mileage, the question from an air quality perspective is how well these vehicles maintain their new vehicle emission levels and, when Received: Revised: Accepted: Published: 5405

February 10, 2016 April 18, 2016 April 19, 2016 May 3, 2016 DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology Table 1. Fleet Sampling Specifics by Location with Measured Emission Means and Standard Errors of the Mean mean (g/kg) of fuel emissions and standard errors of the meana location dates sampled

attempts/plates/matched mean model year

CO

c

NO

/NO2d/NOx

NH3

1.3 ± 0.2

1.5 ± 0.1/−0.04 ± 0.02/2.2 ± 0.1

0.71 ± 0.02

16.4 ± 0.6

2.2 ± 0.2

2.1 ± 0.1/0.15 ± 0.02/3.4 ± 0.1

0.58 ± 0.02

26.3 ± 0.4

6.5 ± 0.3

7.9 ± 0.2/0.24 ± 0.01/12.3 ± 0.2

0.90 ± 0.02

Chicago, IL 9/8−9/13, 2014

26824/20638/20395 2007.5

9.4 ± 0.8

Los Angeles, CA 4/27−5/4, 2013

33807/27808/27184 2004.7 3037/1913/1816 2008.0

SeaTac Int. Airport 9/28−9/29, 2013

HC

b

a

mean speed (mph) acceleration (mph/s) VSPe (kW/tonne) 24.0 0.2 4.8 21.9 −0.2 4.6 15.6 1.2 7.7

a Calculated using a carbon mass fraction of 0.86 for gasoline and diesel, 0.82 for propane, and 0.75 for natural gas. bHC grams expressed using a NDIR correction factor of 2 and normalized as described. cGrams of NO. dGrams of NO2. eVehicle specific power.

aligned across a single-lane road. Collinear beams of infrared (IR) and ultraviolet (UV) light are passed across the roadway into the IR detection unit and are then focused through a dichroic beam splitter that separates the beam into its IR and UV components. The IR detector consists of four nondispersive infrared (NDIR) detectors for a reference channel (3.9 μm), carbon monoxide (CO, 3.6 μm), carbon dioxide (CO2, 4.3 μm), and hydrocarbons (HC, 3.3 μm). The UV beam is focused onto a quartz fiber that divides the light between two fiber bundles and directs the incoming light to two dispersive UV spectrometers. One spectrometer measures nitric oxide (NO), sulfur dioxide (SO2), and ammonia (NH3) between 195 and 225 nm. The second records nitrogen dioxide (NO2) spectra between 430 and 450 nm. All of the detectors sample at 100 Hz and have been described in detail in the literature.12−14 The detector and source are installed on the shoulders of a single-lane roadway with a beam height of approximately 30− 35 cm. When a vehicle blocks the beam, as indicated by a loss of signal on the IR reference channel, the system waits until the beam is reestablished and then collects 0.5 s of data from each IR detector and each spectrometer channel. IR voltages and UV spectra are converted into “concentrations” using an assumed plume path length of 8 cm, which was used initially in the laboratory to generate the detector species response curves. Because the real path length of the plume on-road is unknown, FEAT measures vehicle exhaust gases as a molar ratio to exhaust CO2. Each species data points collected during the 0.5 s sample are plotted against the CO2 data points, and a leastsquares best fit slope (i.e., CO/CO2, HC/CO2, NO/CO2, etc.) is determined. These ratios are constant for a given exhaust plume. The measured ratio of each species is scaled by its certified gas cylinder ratios, which are measured multiple times each day at each location by FEAT to correct for variations in instrument sensitivity and in ambient CO2 levels caused by atmospheric pressure, temperature, and ambient pollution differences. Three similar specification calibration cylinders were used at each site containing (a) 6% CO, 0.6% propane, 6% CO2, and 0.3% NO, balance nitrogen; (b) 0.05% NO2 and 15% CO2, balance air; and (c) 0.1% NH3 and 0.6% propane, balance nitrogen (certified accuracies of ±2%, Air Liquide, Longmont, CO, and Praxair, Tacoma, WA). FEAT has been demonstrated in double-blind intercomparisons to be accurate to ±5% for CO and ±15% for HC for an individual vehicle measurement.15,16 The NO spectrometer has

problems arise, how promptly are the vehicles repaired. The Georgia Institute of Technology and the U.S. EPA collected onroad emission measurements of CO and HC from taxi cabs operating at the Atlanta and New York airports in 1993 and 1994.7 Their results showed that the taxi fleets of both cities had higher average emissions than a sample of typical on-road vehicles. Atlanta taxis were 3 and 36 times higher and New York taxis were 1.5 and 14 times higher for CO and HC, respectively. A similar measurement campaign at the Los Angeles airport in 1992 found similar elevated emissions, and a further investigation exposed an illegal Smog Check certificate generation operation that was successfully prosecuted by the Los Angeles District Attorney’s office.8 Livery fleets have changed significantly in the past decade, with many city fleets now featuring large numbers of hybrid and alternatively fueled vehicles. There has been little recent research into the emission durability of these new fleets, and in this paper, we examine some recent and historical emission data from the Seattle and Chicago area to obtain a picture of how current livery vehicle emissions compare to the rest of the fleet and examine their importance to fleet emissions.



EXPERIMENTAL SECTION This study uses data collected at three U.S. sampling sites (1) in Washington State on the taxi and shuttle loop at the Seattle− Tacoma International Airport (SeaTac, 0° grade), (2) a longterm light-duty site in California located on I-10 between Santa Monica and downtown Los Angeles, CA (WLA, SB La Brea Avenue to EB I-10, 2° grade), and (3) a long-term light-duty site in Illinois located in the northwest suburbs of Chicago (Algonquin Road to SB I-290/SH53, 1° grade), where we have collected emission data since 1997. The SeaTac data were collected on Sept 28 and 29, 2013, as a component of one of the author’s senior thesis; the west LA data were collected over 7 days from April 27 to May 4, 2013; and the 2014 Chicago data were collected over parts of six days from Sept 8−13, 2014. The Chicago historical data sets (1997−2000, 2002, 2004, and 2006) have been previously discussed in the literature, and all of these data sets are available for download from our website at www.feat.biochem.du.edu.9−11 Two identical remote vehicle exhaust sensors, developed by the University of Denver, named Fuel Efficiency Automobile Test (FEAT), were used to collect the emission measurements (serial 3002 in Chicago and LA and serial 3005 in Seattle). The basic instrument is composed of a source and detector unit 5406

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology

vehicles for taxi services at the airport.20 The airport taxis are operated by a single company, and we collected measurements on about 60% of its Prius fleet. Figure 1 plots the mean fuel-

been determined separately to have a minimum detection limit (3σ) of 24 ppm for NO using a new low-emitting test vehicle.13 A comparison of fleet average emission by model year versus IM240 fleet average emissions by model year shows excellent correlations, with R2 values between 0.75 and 0.98, for data from Denver, Phoenix, and Chicago.17 A video image of the license plate of each vehicle is recorded, and the transcribed plate is used to obtain non-personal vehicle information, including make, model year, vehicle identification number, and fuel type from the state registration records of California, Illinois, and Washington. Speed and acceleration measurements for each vehicle are attempted using a pair of parallel infrared beams (Banner Engineering) 1.8 m apart and approximately 0.66 m above the roadway. For this analysis, the measured ratios have been converted into fuel-specific emissions of grams of pollutant per kilogram of fuel by carbon balance using a carbon mass fraction for the fuel of 0.86 and doubling the HC/CO2 ratio to account for the poor quantification of certain hydrocarbon species by NDIR absorption.12,18 Uncertainties presented in the tables and graphs are standard errors of the mean calculated from the distribution of the daily measurement means of each site. Because of only 2 days of data collection in Seattle, uncertainties were estimated using a bootstrap resampling technique, which was used to generate multi-day means from which standard errors of the mean could be calculated (see the Supporting Information).



RESULTS AND DISCUSSION Table 1 contains a summary of the sampling dates, vehicle record information, fleet averaged model year, mean fuelspecific emission measurements with their standard errors of the mean for all of the records, and driving mode parameters collected at each site. Vehicle-specific power has been calculated using the equation from Jimenez et al.19 HC means have been normalized to the lowest HC emitting subfleet at each site for comparison by eliminating site-specific systematic offsets that arise from either liquid water interferences (Chicago, 2014) or instrument-created offsets (see the Supporting Information). SeaTac Airport Fleet. The SeaTac International Airport has a vehicle service fleet composed of many advanced fuel vehicles, which are not commonly found in high enough numbers in the general fleet to study. Table S2 of the Supporting Information lists the number of measurements (1816) and number of unique vehicles (524) by fuel type for the SeaTac database. Because of the nature of their business, there are many repeat measurements on the same vehicle. The measurements were collected on the flat covered taxi and shuttle bus loop road exit that parallels the terminal. Only taxis, limousines, and shuttle buses are permitted on this road. The shuttle buses were constantly circulating, stopping only long enough to pick up or drop off passengers. Taxi driving modes are less predictable, with longer stops a distinct possibility, although each taxi had to have driven more than a quarter of a mile before reaching our measurement location. Measurements were collected between 10:30 and 19:00 h over the 2 days with temperatures between 12 and 17 °C. Slightly more than a quarter of the measurements (506; see Table S2 of the Supporting Information) were collected on Toyota Prius models, whose fleet fractions have increased at the airport as a result of a relaxed 2006 city rule, which allowed gasoline hybrid vehicles to substitute for natural-gas-fueled

Figure 1. SeaTac Toyota Prius taxis mean grams of pollutant per kilogram of fuel emissions (●) by model year compared to Toyota Prius model emissions measured on-road at the west Los Angeles measurement site (△). Uncertainties are standard error of the means determined from the daily samples for the west LA data and by bootstrap resampling techniques for the SeaTac measurements.

specific emissions by model year for CO (top), HC (middle), and NOx (bottom) for the Prius taxis measured at SeaTac (●) and Prius vehicles measured a few months earlier at the WLA location (△). Prius measurements make up approximately 2.3% of the 2013 WLA database (619/27247 with 509 unique vehicles), but only 24 measurements are from taxis. Concerns about whether the airport taxis have engines and catalytic converters at operating temperature are allayed by the fact that, if our airport measurements had a large fraction of cold-start vehicles, we would expect NOx emissions to be depressed in all model years and CO and HC emissions to be similarly increased when compared to the WLA site, where there is little possibility for cold-start vehicles. Figure 1 shows that CO and HC emissions are statistically similar for the two data sets for the first 4 model years, after which the SeaTac Prius emission values rise rapidly. NOx emissions for the airport taxis are higher during the first 4 model years than the WLA vehicles but also rise to higher levels starting with the 2009 models. The SeaTac Prius NH3 emissions also show an increase (see Figure S1 of the Supporting Information) after the first 4 model years. The drop in all species emissions with the 2006 SeaTac models may or may not be representative because there were only four measurements from two vehicles collected for that model year. These observed emission differences between the Prius vehicles measured at SeaTac and in WLA are likely connected with how the two fleets are used and maintained. At the time of 5407

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology

emissions versus model year for the gasoline (●) and propane (△) fueled SeaTac shuttle buses. NH3 emissions for the shuttle buses peak between 5 and 6 year old vehicles and then begin to decline, indicating that, on average, the catalysts of these vehicles begin to lose their reduction capability about 3−4 times sooner than the on-road WLA fleet. This is another indication that these vehicles accumulate mileage at a higher rate than the typical on-road fleet and that exhaust aftertreatment components degrade as a result. Chicago Fleet. At our long-term suburban data collection site northwest of Chicago, we have collected emission data sets in the years 1997−2000, 2002, 2004, 2006, and most recently in 2014.11 In Illinois, traditional taxi and limousines have dedicated license plate formats (taxi plates end in TX, and limousine plates end in LY) that allow them to be easily distinguished from other vehicles in the databases. Using this information, we analyzed and compared the emissions of the taxi and limousine fleets to the general fleet emission distributions and patterns for each measurement year. Table 2 provides a summary of the number of measurements for each fleet and mean model year by measurement year and shows that the taxi fleet of this site is consistently older than the onroad fleet, while the limousine segment is consistently younger. Figures 3 and 4 plot fuel-specific emissions for CO, HC, and NO emissions for the taxi and limousine fleets and age-adjusted fleet emissions for the data set of each year. Measurements of NO2 emissions are only available for the 2014 data set; therefore, we have chosen to make all of the comparisons using NO. Because age is the most significant determining factor in fleet emission means, it is important that we eliminate any age differences before making the comparison between the Chicago sites on-road fleet and the taxi and limousine fleets. This is accomplished by using the age distribution of taxi or limousine fleets and applying it to the emission distribution for the onroad fleet, producing an on-road fleet emission mean that now has the same age as the taxi or limousine fleet. This process is detailed in Table S3 of the Supporting Information. Mean taxi emissions at our northwest Chicago site are consistently higher (Figure 3) than for the similarly ageadjusted on-road fleet for all three pollutants. The one exception is for the HC emissions in the 1998 data set, which is also abnormally low for the limousines. The process of correcting for fleet zero emission offsets in each year’s data set (see the Supporting Information) has overcorrected for the 1998 taxi and limousine subfleets. The higher emissions indicate that the taxis have a higher proportion of vehicles in need of emission-related maintenance than the general fleet.

these measurements, emission inspections were not required in Seattle nor California for any Prius models; however, Seattle taxis are restricted to being only 7 years old or newer, and beginning in the spring of 2015, California brought 7 year and older Prius models into their onboard diagnostic-only testing program.21 One major difference between the two fleets is that we expect the rate at which mileage is accumulated on the airport taxis to be significantly higher than the largely public owned WLA vehicles. Anecdotal data from talking with some of the SeaTac taxi drivers was that 200 miles/day was common. This would quickly put 50 000 miles or more on a taxi in a year’s time, which is about 4 times the national average mileage accumulation rate of 13 500 miles/year.22 One way to gauge mileage accumulation rates is through its negative effects on three-way catalyst activity levels. It is wellknown that catalyst age and mileage accumulation is linked and catalysts deteriorate with age.23,24 We can gauge three-way catalyst activity levels for non-diesel-fueled vehicles by their ability to catalyze the reduction reaction of NO to NH3. We have previously reported that fleet NH3 emissions rise with vehicle age until the catalyst begin to lose their ability to reduce NO, at which point NH3 emissions decline back to the low levels observed in vehicles not equipped with three-way catalytic converters.4,10,25,26 In 2013 at the WLA site, the NH3 emission turnover point occurred between 16 and 19 year old vehicles. Figure 2 plots grams of NH3 per kilogram of fuel

Figure 2. Grams of NH3 per kilogram of fuel emissions plotted versus model year for gasoline (●) and propane (△) fueled shuttle vans measured at SeaTac International Airport. Uncertainties plotted are standard errors of the mean determined using bootstrap resampling techniques.

Table 2. Chicago Taxi and Limousine Vehicle Fleet Statistics, Fleet Percentages, and Percent Emission Contribution percent emission contribution taxis/limo (multiple of fleet percentage) year

measurements of fleet/taxis/limo mean model year (fleet/taxis/limo)

2014 20395/268/183 (2007.5/2007.3/2011.3) 2006 22200/341/106 (2001/2000/2002.8) 2004 21838/141/70 (1999.2/1997.7/2000.9) 2002 22320/160/104 (1997.4/1994.9/1999) 2000 22065/154/106 (1995.5/1993.1/1997.1) 1999 23088/146/182 (1994.6/1992.6/1996.4) 1998 23560/84/151 (1993.6/1992.5/1995.4) 1997 19682/132/132 (1992.7/1991.1/1994.2) vehicle weighted means

fleet percentages of taxis/limo 1.3/0.9 1.5/0.5 0.6/0.3 0.7/0.5 0.7/0.5 0.6/0.8 0.4/0.6 0.7/0.7 1.0/0.6 5408

CO 3.2/0.6 2.7/0.5 1.2/0.5 1.8/0.6 1.6/0.5 1.8/0.9 0.6/0.5 1.5/0.7 2.1/0.6

(2.4/0.7) (1.8/1) (1.9/1.4) (2.5/1.2) (2.3/1.0) (2.8/1.1) (1.8/0.8) (2.2/1.0) (2.1/1.0)

HC 3.1/0.6 (2.4/0.6) 3.2/0.6 (2.1/1.3) 1.6/0.4 (2.5/1.3) 2.1/0.5 (2.9/1.1) 1.7/0.3 (2.4/0.6) 2.0/0.5 (3.2/0.6) 0.1/0 (0.2/0) 2.4/0.4 (2.4/0.6) 2.4/0.4 (2.4/0.6)

NO 2.4/0.3 2.4/0.3 1.5/0.4 1.8/0.6 1.4/0.5 1.2/1.0 0.5/0.7 0.9/0.9 1.8/0.6

(1.8/0.4) (1.6/0.3) (2.3/1.2) (2.5/1.3) (2.0/1.0) (1.9/1.2) (1.4/1.1) (1.8/1.3) (1.8/1.0)

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology

Figure 3. Chicago historical mean grams of pollutant per kilogram of fuel emissions by measurement year for taxis (○) and the entire Chicago fleet that has been age-corrected to match the age of the taxi subfleet (●) for each data set collected at this site. Uncertainties plotted are standard error of the means determined using the daily samples.

Figure 4. Chicago historical mean grams of pollutant per kilogram of fuel emissions by measurement year for limousines (○) and the entire Chicago fleet that has been age-corrected to match the age of the limousine subfleet (●) for each data set collected at this site. Uncertainties plotted are standard error of the means determined using the daily samples.

The good news is that taxi emissions have decreased along with the general fleet, with large reductions in emissions over the 17 year span of measurements. The historical limousine emission comparison (Figure 4) shows that emissions have also declined over the 17 year period and that CO and NO emissions are not statistically different from the general fleet in 2006 and for all three species in 2014. Table 2 lists the number of measurements and fleet percentage representation for the taxis and limousines for each measurement year and the percentage of the total fleet emissions that the taxi and limousine fleets contributed. This is not an age-normalized analysis but a check of whether these Chicago subfleets (taxi and limousine) emission contributions are proportional to their fleet representation. For example, the 2014 taxi fleet is responsible for 3.2, 3.1, and 2.4% of the fleet CO, HC, and NO emissions, respectively, even though they only comprise 1.3% of the measured fleet. These percent contributions assume that, on average, the taxis consume the same amount of fuel as the rest of the fleet. If, as we expect, the taxis use 2−4 times more fuel in their travels, then the emission contribution percentages increase by these factors. Overall, the taxi fleet is responsible for, on average, 2.1, 2.4, and 1.8 times more emissions for CO, HC, and NO, respectively, over the past 17 years than would be expected from their fleet percentages, indicating a disproportionate emissions share for the taxis. The limousine fleet over this same time period has

averages of 1, 0.6, and 1 for CO, HC, and NO, respectively, indicating that the limousine emissions are generally inline or a little lower than their numerical representation. The differences between emission contributions are also born out when we compare the two 2014 subfleet emission distributions with the 2014 on-road fleet. Figure 5 is a percentile−percentile plot of the taxi fleet’s fuel-specific emissions for CO, HC, and NO compared to the on-road fleet. The diagonal line is the 1:1 line showing where the points would fall if the emission distributions of the two fleets were identical. For both the CO and NO emissions, the taxi fleet starts to have higher emissions than the on-road fleet at very low percentiles. For HC, the 2014 taxi fleet has generally slightly higher emissions across all percentiles until the 99th percentile, where the deviation is more pronounced. The taxis have significantly higher 99th percentile emissions, a percentile that we consider to be fully populated with broken vehicles for all three pollutants. These differences are not likely just the result of age because the 2014 taxi fleet is only 0.2 of a model year older than the on-road Chicago fleet, indicating that other factors are involved in the emission differences seen. A similar plot (Figure S2 of the Supporting Information) compares the 2014 limousine subfleet to the on-road fleet and shows that, for CO and HC, the emission distribution follows the 1:1 line, while NO emissions are significantly lower, which may be a result of the age differences (limousine fleet is 3.8 years 5409

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology

CO, HC, and NOx emissions per kilogram of fuel by model year for the 2014 Chicago area taxis (all makes and models) and limousine fleets and compares them against the 2014 Chicago on-road fleet. The 2014 taxi (268 measurements) and limousine (183 measurements) subfleets measured in Chicago have fewer measurements, which leads to larger uncertainties in the fuel-specific emission values when broken out by model year. While the 2014 Chicago area taxi fleet in general has higher emissions than the general fleet and the 2014 limousine fleet has similar or lower emissions as previously discussed, there is no clear change in emission levels with increasing age, as observed with the SeaTac airport fleet. Fleet emission inventory calculations generally rely on some type of age-weighted activity factor, usually model year, for generating the totals. High-mileage vehicles, such as taxis and limousines, are not generally considered to be numerous enough to change these weighting factors, because, for example, in King County, Washington, taxis only account for about 0.6% of the registered vehicles (2000/325 000). If, however, as observed in Chicago, their emission contribution is about twice their fleet percentage and, in addition, they consume more fuel than the county average, then their emission contribution could grow to be 2−5% of the total. In areas where these highmileage fleets are in higher concentrations, such as airports and downtown business centers, their emission contribution could produce much larger inventory errors if not taken into account. As tailpipe emissions continue to decrease, emission distributions will become more skewed, likely increasing the importance of higher emission high-mileage fleets and their contribution to vehicle emission inventories.



ASSOCIATED CONTENT

S Supporting Information *

Figure 5. Percentile−percentile plots of the 2014 emission distribution comparison of the Chicago site taxi fleet versus the on-road fleet of fuel-specific emissions for CO (top), HC (middle), and NO (bottom). Lines in each plot are 1:1 lines drawn to indicate a perfect match between the two emission distributions.

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.6b00717.



younger than the on-road fleet and 4 years younger than the taxi fleet) between these two groups. Even when corrected for age differences, the 2014 taxi fleet still has significantly higher emissions for CO, HC, and NO than the 2014 limousine fleet. Table S4 of the Supporting Information details the percentage of makes and mean emissions for the taxi and limousine fleets for the 2014 Chicago data set. The limousines are dominated by Lincolns (115/183) and Cadillacs (28/183), with these two groups representing 78% of all of the measurements. These two makes have significantly lower emissions than most of the taxi makes and are the major reason that the limousines have lower overall emissions. The taxi fleet is more diverse, with the two largest makes represented being Toyota (86/268, of which 31 are Prius models) and Dodge (65/268), accounting for 56% of the taxis measured. The emission differences likely come down to the differences in maintenance practices between the two fleets, but without specific information to that effect, we cannot speculate. The 4 year old Toyota Prius taxis measured at SeaTac showed a pattern of increasing emissions (see Figure 1) when compared to similarly aged models measured in Los Angeles. To check for similar emissions behavior for the Chicago taxis, Figure S3 of the Supporting Information plots the grams of

Supplementary tables (Tables S1−S4) and figures (Figures S1−S3) referenced in the text (PDF)

AUTHOR INFORMATION

Corresponding Author

*Telephone: 303-871-2584. E-mail: [email protected]. Notes

The authors declare the following competing financial interest(s): The authors Gary A. Bishop and Donald H. Stedman acknowledge receipt from the University of Denver of a share of the patent royalty payments from Envirotest, an operating subsidiary of Opus Inspection, which licenses the commercial remote sensing technology developed at the University of Denver. Commercial instrumentation was not used in this study.



ACKNOWLEDGMENTS

This paper is a result of work supported by the California Environmental Protection Agency Air Resources Board (12303), the Coordinating Research Council (E-106), the University of Puget Sound, and the University of Denver. Results and conclusions presented here are solely the responsibility of the authors and may not represent the views of the sponsors. The authors gratefully acknowledge Annette Bishop (license plate transcription). 5410

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411

Article

Environmental Science & Technology



(19) Jimenez, J. L.; McClintock, P.; McRae, G. J.; Nelson, D. D.; Zahniser, M. S. Vehicle specific power: A useful parameter for remote sensing and emission studies. Proceedings of the Ninth CRC On-Road Vehicle Emissions Workshop; Coordinating Research Council: Alpharetta, GA, 1999; pp 7-45−7-57. (20) Murakami, K. First hybrid taxis humming at Sea-Tac. Seattle Post-Intelligencer Sept 13, 2007; http://www.seattlepi.com/local/ article/First-hybrid-taxis-humming-at-Sea-Tac-1249560.php (accessed Feb 2016). (21) California Bureau of Automotive Repair. Q&As: Hybrid Vehicles in the Smog Check Program; http://www.bar.ca.gov/Industry/ Industry_Resources/Q&As_Hybrid_Vehicles_in_the_Smog_Check_ Program.html#Question1 (accessed April 2016). (22) Federal Highway Administration, United States Department of Transportation. Average Annual Miles Per Driver by Age Group; https:// www.fhwa.dot.gov/ohim/onh00/bar8.htm (accessed April 2016). (23) López Granados, M.; Larese, C.; Cabello Galisteo, F.; Mariscal, R.; Fierro, J. L. G.; Fernández-Ruíz, R.; Sanguino, R.; Luna, M. Effect of mileage on the deactivation of vehicle-aged three-way catalysts. Catal. Today 2005, 107−108, 77−85. (24) Winkler, A.; Ferri, D.; Hauert, R. Influence of aging effects on the conversion efficiency of automotive exhaust gas catalysts. Catal. Today 2010, 155 (155), 140−146. (25) Bishop, G. A.; Peddle, A. M.; Stedman, D. H.; Zhan, T. On-road emission measurements of reactive nitrogen compounds from three California cities. Environ. Sci. Technol. 2010, 44, 3616−3620. (26) Burgard, D. A.; Bishop, G. A.; Stedman, D. H. Remote sensing of ammonia and sulfur dioxide from on-road light duty vehicles. Environ. Sci. Technol. 2006, 40, 7018−7022.

REFERENCES

(1) Air Resources Board, California Environmental Protection Agency. EMFAC Emissions Database; http://www.arb.ca.gov/emfac/ (accessed Feb 2016). (2) United States Environmental Protection Agency. Modeling and Inventories; MOVES (Motor Vehicle Emission Simulator); http://www. epa.gov/otaq/models/moves/ (accessed Feb 2016). (3) Zhang, Y.; Bishop, G. A.; Stedman, D. H. Automobile emissions are statistically gamma distributed. Environ. Sci. Technol. 1994, 28, 1370−1374. (4) Bishop, G. A.; Schuchmann, B. G.; Stedman, D. H.; Lawson, D. R. Multispecies remote sensing measurements of vehicle emissions on Sherman Way in Van Nuys, California. J. Air Waste Manage. Assoc. 2012, 62 (10), 1127−1133. (5) Bishop, G. A.; Stedman, D. H. On-Road Remote Sensing of Automobile Emissions in the Tulsa Area: Fall 2013; Coordinating Research Council: Alpharetta, GA, 2014; http://www.feat.biochem.du. edu/assets/databases/Oklahoma/CRC106%20Tulsa%202013%20Final%20report.pdf (accessed Feb 2016). (6) Berk, B. Taxi tough: A look inside an NYC cab shop. Car Driver Feb 2011; http://www.caranddriver.com/features/taxi-tough-a-lookinside-an-nyc-cab-shop-feature. (7) Vescio, N.; DuBose, R. S.; Rodgers, M. O. Comparison of taxi fleet compositions and emissions behavior in Atlanta, Georgia, and New York City. Proceedings of the 5th Coordinating Research Council’s On-Road Vehicle Emission Workshop; San Diego, CA, April 3−5, 1995. (8) Churchill, M. Auto smog program needs repairs. Press Enterprise July 26, 1992. (9) Bishop, G. A.; Stedman, D. H. Reactive Nitrogen Species Emission Trends in Three Light-/Medium-Duty United States Fleets. Environ. Sci. Technol. 2015, 49 (18), 11234−11240. (10) Bishop, G. A.; Stedman, D. H. The recession of 2008 and it impact on light-duty vehicle emissions in three western US cities. Environ. Sci. Technol. 2014, 48, 14822−14827. (11) Bishop, G. A.; Stedman, D. H. A decade of on-road emissions measurements. Environ. Sci. Technol. 2008, 42 (5), 1651−1656. (12) Burgard, D. A.; Bishop, G. A.; Stadtmuller, R. S.; Dalton, T. R.; Stedman, D. H. Spectroscopy applied to on-road mobile source emissions. Appl. Spectrosc. 2006, 60, 135−148. (13) Popp, P. J.; Bishop, G. A.; Stedman, D. H. Development of a high-speed ultraviolet spectrometer for remote sensing of mobile source nitric oxide emissions. J. Air Waste Manage. Assoc. 1999, 49, 1463−1468. (14) Burgard, D. A.; Dalton, T. R.; Bishop, G. A.; Starkey, J. R.; Stedman, D. H. Nitrogen dioxide, sulfur dioxide, and ammonia detector for remote sensing of vehicle emissions. Rev. Sci. Instrum. 2006, 77, 014101. (15) Lawson, D. R.; Groblicki, P. J.; Stedman, D. H.; Bishop, G. A.; Guenther, P. L. Emissions from in-use motor vehicles in Los Angeles: A pilot study of remote sensing and the inspection and maintenance program. J. Air Waste Manage. Assoc. 1990, 40, 1096−1105. (16) Ashbaugh, L. L.; Lawson, D. R.; Bishop, G. A.; Guenther, P. L.; Stedman, D. H.; Stephens, R. D.; Groblicki, P. J.; Johnson, B. J.; Huang, S. C. On-road remote sensing of carbon monoxide and hydrocarbon emissions during several vehicle operating conditions. Proceedings of the A&WMA International Specialty Conference on PM10 Standards and Non-traditional Source Control; Phoenix, AZ, Jan 12−15, 1992. (17) Pokharel, S. S.; Stedman, D. H.; Bishop, G. A. RSD Versus IM240 Fleet Average Correlations. Proceedings of the 10th CRC OnRoad Vehicle Emissions Workshop; San Diego, CA, March 27−29, 2000; http://www.feat.biochem.du.edu/assets/presentations/ RSD%20verus%20IM240%20Correlations%2010th_CRC_2000.pdf (accessed Feb 2016). (18) Singer, B. C.; Harley, R. A.; Littlejohn, D.; Ho, J.; Vo, T. Scaling of infrared remote sensor hydrocarbon measurements for motor vehicle emission inventory calculations. Environ. Sci. Technol. 1998, 32, 3241−3248. 5411

DOI: 10.1021/acs.est.6b00717 Environ. Sci. Technol. 2016, 50, 5405−5411